AI-based service generates a realistic approximation of a person based on old photographs, paintings, drawings, sketches, and other media.
Check out a new update of a neural network that generates realistic HD portraits from artistic or photographic sources. The process is straightforward: you just prepare an image and then upload it for the service to study. Then, you get an image of how a subject from the artwork would look in real life. When the portrait is generated, users can either compare it with the original artwork on the software or receive a result by email. neural.love notes that generated content is not historically accurate, however, it is fun to recreate the faces of people from the past.
Van Gogh
Anne Frank
neural.love was launched in 2020 and since then, it kept improving its multiple neural networks. The updated version of the AI-powered HD portrait generator allows achieving a realistic approximation of a person based on photographs, paintings, drawings, sketches, and other media. According to neural.love, the Portrait Generator tool has become more accurate compared to the previous versions. However, the company warned users that the speed has slowed a bit, so it'll take a little more time now to generate the portrait.
Margaret Thatcher
Fyodor Dostoevsky
neural.love also informed that the tool may not work on modern digital photos, sources with visible texture, damage, or contrast issues, as well as subjects with obscured or distorted faces, complex hairstyles or hats, and unique facial characteristics.
Find out more about the project here. Also, don't forget to join our new Reddit page, our new Telegram channel, follow us on Instagram and Twitter, where we are sharing breakdowns, the latest news, awesome artworks, and more.